Letters in Applied Microbiology 2001, 32, 303±306
Comparative study of the in¯uence of melatonin
and vitamin E on the surface characteristics
of Escherichia coli
J. Uberos1, C. Augustin1, J. LieÂbana2, A. Molina1 and A. MunÄoz-Hoyos1
1
Research Group CTS-190, Department of Paediatrics, `San Cecilio' University Clinical Hospital, and
Department of Microbiology, School of Medicine, Granada, Spain
2
2000/58: received 29 January 2001 and accepted 9 February 2001
Ä O Z - H O Y O S . 2001.
J . U B E R O S , C . A U G U S T I N , J . L I EÂ B A N A , A . M O L I N A A N D A . M U N
2
Aims: Melatonin is a hormone produced by the pineal gland and that affects the response of
various cell membranes to an oxidative stimulus.
Methods and Results: The present study evaluates the hydrophobic characteristics of
Escherichia coli in response to melatonin (100 nmol l±1, 200 lmol l±1) and to vitamin E
(5 mg dl)1). A reduction was found in the surface hydrophobicity of E. coli at concentrations of
200 lmol l±1 melatonin in a MuÈeller-Hinton (MH) broth. These effects were modi®ed when a
protein synthesis inhibitor (chloramphenicol) was added at sub-lethal concentrations to the
broth. Vitamin E produced a greater diminution in surface hydrophobicity than melatonin.
The adherence of E. coli to nitrocellulose ®lters increased in the presence of melatonin + chloramphenicol, and vitamin E. The effects observed were independent of the
concentration of iron in the broth.
Conclusions: Oxidative stress plays an important role in modifying the surface characteristics
of E. coli, which could affect the micro-organism's capacity to adhere to epithelia.
Signi®cance and Impact of the Study: We think that the oxide reduction potential of the
host may be a determinant factor in the bacterial colonization of animal tissue.
INTRODUCTION
The ®rst stage in the colonization of an epithelium by
Escherichia coli is determined by the bacteria's ability to
adhere to the epithelial surface, a capability that is initially
dependent on non-speci®c adhesion mechanisms such as the
electrical charge at the surface of the micro-organism or the
surface hydrophobicity (FerreiroÂs and Criado 1984). Other
speci®c factors, such as the formation of pili or speci®c
adhesins, subsequently contribute to the adhesion of the
micro-organism to epithelia (FerreiroÂs and Criado 1980;
Gibbons 1996). The relative importance of speci®c or nonspeci®c adhesion mechanisms varies between micro-organisms, and also depends on their adaptation to the colonization of speci®c epithelia (Gibbons 1996). Melatonin, which
is a hormone found in the pineal gland, plays a stabilizing
and regulatory role in humans and is also signi®cant in sleep
Correspondence to: Dr Jose Uberos, Dpto. PaediatrõÂa, Hospital Universitario,
Granada, Spain (e-mail: med018111@nacom.es).
ã 2001 The Society for Applied Microbiology
regulation, the immune response, the elimination of free
radicals and the stabilization of cell membranes (Li and Xu
1997; RodrõÂguez et al. 1997; Reiter et al. 1997).
The present study evaluates possible relations between
melatonin, vitamin E and the surface hydrophobic characteristics of E. coli. In addition, as iron has been shown to be an
important factor in the virulence of many micro-organisms
(Criado et al. 1993; Stojiljkovic et al. 1995), it was determined whether the different concentrations of iron in the
culture medium, in conjunction with melatonin or vitamin E,
modify the surface characteristics of this micro-organism.
The initial focus of the study is the hypothesis that the
oxide reduction potential of the host may be a determinant
factor in the bacterial colonization of animal tissue.
MATERIALS AND METHODS
Bacterial strains and culture conditions
Eleven strains of E. coli were obtained from the Spanish
Collection of Standard Cultures (CECT; Valencia, Spain)
304 J . U B E R O S E T A L .
(ATCC 25922, ATCC 11775, Rowe E8775-A, CCUG
20570, CGSC 6512, ATCC 23985, ATCC 8739, ATCC
12407, NCIMB 9482, ATCC 33780, Starlinger 1045).
Twelve strains of E. Coli (Lac+), isolated from upper
pathway urinary infections in the microbiology laboratory at
this Hospital (629, 695, 787, 472, 471, 697, 595, 753, 795,
792, 760, 469), were also used.
In each test, E. coli was cultivated in 4 ml MuÈeller-Hinton
(MH) broth (Fluka) and incubated at 36 ‹ 1°C for 24 h
before being processed. The iron content of the culture
medium was modi®ed by adding 39 lmol l±1 ethylenediamine-dihydroxiphenyl-acetic acid (Sigma) to 100 ml MH
(EDDA-MH, divalent cations restriction), or by adding
ferric-ammonium-citrate (Sigma) (1 lg ml±1) (FAC-MH,
iron excess).
Solutions of melatonin (Sigma; 1 mmol l±1 and
500 nmol l±1) were prepared in sterile distilled water. After
incubating the strains of E. coli for 12 h at 36 ‹ 1°C
(logarithmic growth phase), 1 ml of each of the solutions was
added to each tube of the culture medium (4 ml), resulting
in a ®nal concentration of 200 lmol l±1 and 100 nmol l±1.
Incubation was continued for 12 h at 36 ‹ 1°C until the
process was complete.
Surface hydrophobicity
The surface hydrophobicity was determined by Rosenberg's
method (Rosenberg et al. 1980). The organisms were
incubated at 36 ‹ 1°C for 24 h in a culture medium and
then recovered by centrifugation. The supernatant ¯uid was
removed and the sediment washed three times with saline
phosphate buffer (PBS). The sediment was centrifuged
again and then mixed with 0á4 ml urea buffer tampon
(PUM) (pH 7á1) and 0á8 ml 0á2 mol l±1 ammonium sulphate
(Sigma). The suspension was adjusted to an optical density
(O.D. 560 nm) of 0á6; 200 ll p-xylene (Sigma) were added
to 1á2 ml of the bacterial suspension, which was then stirred
vigorously for 2 min to obtain a homogeneous mixture of the
suspension and the hydrocarbon. This was allowed to stand
for 30 min in order for the aqueous phase to separate
completely from the hydrocarbon. Using a glass Pasteur
pipette, 1 ml of the aqueous phase was then removed and
the absorbance measured (560 nm) at a temperature of
36 ‹ 1°C. Hydrophobicity was calculated as 1 ) (A2/A1),
where A1 and A2 represent the absorbance before and after
the hydrocarbon was added, respectively.
Adherence to nitrocellulose ®lters
The procedure described by Lachica and Zink was used
(Lachica and Zink 1984). The organisms were placed in a
culture broth for 24 h at 36 ‹ 1°C and centrifuged at 3500 g
for 10 min. The supernatant ¯uid was removed and the
sediment washed three times with phosphate buffer (PBS).
The micro-organisms were resuspended in PUM buffer and
ammonium sulphate (1 : 2, v/v), after the absorbance had
been adjusted to 0á6 at 560 nm. Then, 3 ml of this
suspension were passed through a nitrocellulose ®lter of
8 lm pore diameter (Millipore). The O.D. of the ®ltrate was
determined and the results expressed as a percentage of the
adherence to the nitrocellulose ®lter.
Incubation with vitamin E
The disodium salt (250 mg) of alpha-tocopherol phosphate
(Sigma) was dissolved in 50 ml distilled water. The microorganisms were incubated at 36 ‹ 1°C for 12 h in tubes
containing 4 ml of the MH, EDDA-MH and FAC-MH
media. Then, 40 ll of the vitamin E solution were added,
thus obtaining a concentration of 5 mg dl±1 of vitamin E in
the culture broth. The incubation was continued for a
further 12 h until analysis.
Incubation with chloramphenicol at sub-lethal
concentrations
4 Chloramphenicol at sub-lethal concentrations (100 lg ml±1)
inhibits protein synthesis by micro-organisms (Fu et al.
1989). Before each test, solutions of sodium succinate
chloramphenicol (Sigma) (1 mg ml±1) were prepared. After
incubating the bacteria for 12 h in MH, EDDA-MH and
5 FAC-MH media, 400 ll of the chloramphenicol solution
and 1 ml of the melatonin solution (see above) were added to
each tube of culture medium (4 ml). Incubation was
completed at 36 ‹ 1°C for 12 h.
Statistical methods
A Kolmogorov normality study was carried out. After
observing a lack of normality in the distributions, the
Wilcoxon comparison test for paired data and the Kendall
agreement test were performed.
RESULTS
Table 1 gives the mean values for surface hydrophobicity
and adherence to nitrocellulose ®lters obtained for E. coli
incubated with melatonin 1 mmol l±1, melatonin
1 mmol l)1 + chloramphenicol and vitamin E in MuÈellerHinton broth (MH, normal iron), MH with ferric-ammonium citrate (FAC-MH, 1 g Fe ml±1, iron-excess) and with
20 lmol l±1
ethylene-diamino-dihydroxyphenylacetic
(EDDA-MH, iron restriction). The concentration of iron
in the culture broth had no signi®cant effect on the surface
hydrophobicity, or on the adherence to nitrocellulose ®lters,
of E. coli.
ã 2001 The Society for Applied Microbiology, Letters in Applied Microbiology, 32, 303±306
MELATONIN AND ESCHERICHIA COLI
1
305
Table 1 Cell-surface hydrophobicity and adherence to nitrocellulose ®lters of Escherichia coli incubated with melatonin (aMT), vitamin E and
chloramphenicol + aMT, for a MuÈeller-Hinton broth (MH), for iron excess (FAC-MH) and for iron restriction (EDDA-MH)
% Hydrophobicity (S.D.)
% Adherence to nitrocellulose (S.D.)
MH EDDA-MHà
FAC-MHàà
MH EDDA-MHà
FAC-MHàà
No aMT
16á9
(21á3)
17á6
(19á7)
15á3
(19á6)
81á4
(19á2)
78á9
(24á2)
84á9
(13á7)
aMT (100 nmol l)1)
16á1
(19á0)
17á9
(19á7)
15á5
(19á8)
83á5
(16á5)
79á8
(16á3)
84á1
(16á5)
aMT (200 lmol l)1)
6á5*
(11á6)
10á1
(13á9)
8á5
(12á9)
77á8
(19á1)
71á8
(24á9)
72á9**
(24á7)
aMT (200 lmol l)1)
+ Chloramphenicol
11á2
(19á1)
11á9
(17á2)
12á4
(20á9)
88á6§
(11á7)
83á3
(18á1)
88á6§§
(11á6)
Vitamin E
0á23±±
(1á1)
4á1±
(17á7)
0±±
(0á0)
89á2±
(10á8)
86á3±
(13á6)
90±±
(11á1)
MuÈeller-Hinton broth; àMueller-Hinton broth + ethylene-diamino-dihydroxyphenilacetic acid; ààferric ammonium citrate + Mueller-Hinton
broth.; (*) P < 0á05 (**) P < 0á01, comparison between melatonin 0 mmol l)1 and melatonin 200 lmol l)1; (±) P < 0á05 (±±) P < 0á01, comparison
between melatonin 200 lmol l)1 and vitamin E.; (§) P < 0á05 (§§) P < 0á01, comparison between melatonin 200 lmol l)1 and melatonin
200 lmol l)1 + chloramphenicol.
The surface hydrophobicity of E. coli diminished signi®cantly after incubation with melatonin at a concentration of
200 lmol l±1 in MH broth. There was a similar reduction in
the rate of adherence to nitrocellulose in FAC-MH broth.
The addition of sub-lethal concentrations of chloramphenicol + melatonin (200 lmol l±1) to the broth did not produce
signi®cant changes in surface hydrophobicity. Under these
conditions, the percentage of adherence to nitrocellulose
®lters increased signi®cantly when E. coli was incubated in the
presence of iron, as the addition of a divalent cation chelating
agent (ethylene-diamino-dihydroxyphenylacetic acid) to the
broth reduced the percentage of adherence to nitrocellulose.
Vitamin E produced a signi®cant reduction in the surface
hydrophobicity of E. coli, much greater than that obtained
with melatonin 200 lmol l±1. The percentage of adherence
of E. coli to nitrocellulose ®lters increased signi®cantly when
the micro-organism was incubated in the presence of
vitamin E. In these cases, the changes in surface hydrophobicity and adherence to nitrocellulose were greater when
the culture broth was supplemented with iron.
The Kendall agreement test between adherence to nitrocellulose and surface hydrophobicity only revealed agreement between the two measurement methods (T ˆ 0á52;
P ˆ 0á001) when E. coli was incubated with chloramphenicol
and melatonin (200 lmol l)1) in EDDA-MH.
DISCUSSION
The ®rst stage in the colonization of an epithelium by E. coli
depends on the bacteria's ability to adhere to the cells of the
host; initially, this ability is determined by the electrical
charge at the surface of the micro-organism and by its
surface hydrophobicity (FerreiroÂs and Criado 1984), and
subsequently, by factors such as the formation of pili and
speci®c adhesins (Gibbons et al. 1976; FerreiroÂs and Criado
1980; Gibbons and Hay 1989).
Previous studies (Uberos et al. 2000) have shown that
melatonin at a concentration of 200 lmol l±1 signi®cantly
increases the surface hydrophobicity of Neisseria meningitidis
and its adherence to nitrocellulose, and that the concentration of iron in the culture broth is not a signi®cant factor in
such differences. These conclusions, together with those of
the present study, lead to the belief that the effects of
melatonin on the surface characteristics of micro-organisms
vary between different micro-organisms.
Several studies (Ellis and Vorhies 1976; Hogan et al. 1993;
Friedman et al. 1998) have shown how the administration of
high concentrations of vitamin E to animals is associated
with higher levels of antibodies against E. coli. It is not
known whether this ®nding bears any relation to the reduction in surface hydrophobicity revealed in the present study,
although in theory a reduction in the surface hydrophobicity
of E. coli would impede the colonization of epithelia, both
gastrointestinal and genito-urinary. Furthermore, the adjuvant effect described for vitamin E (Hogan et al. 1993) in the
induction of the humoral immune response could be a
consequence of the altered surface hydrophobicity characteristics of E. coli incubated with vitamin E, namely, the
reduction in surface hydrophobicity and the increase in
the percentage of adherence to nitrocellulose.
Chloramphenicol at inhibitory concentrations inhibits
protein synthesis by micro-organisms (Fu et al. 1989).
ã 2001 The Society for Applied Microbiology, Letters in Applied Microbiology, 32, 303±306
306 J . U B E R O S E T A L .
The present study included incubation with chloramphenicol and melatonin (200 lmol l±1) to test whether the effects
observed on the surface characteristics of E. coli were due to
protein synthesis de novo. It was found that melatonin leads
to a reduction in the surface hydrophobicity of E. coli, a
reduction which is unaltered by incubation with chloramphenicol + melatonin. Therefore, it is concluded that the
effect of melatonin on the surface hydrophobicity of E. coli is
due to the synthesis de novo of surface proteins of the microorganism. When E. coli (in the presence of iron) was
incubated with melatonin + chloramphenicol, the adherence to nitrocellulose ®lters increased signi®cantly to levels
close to those found when vitamin E was included. The
inhibition of protein synthesis by chloramphenicol was very
probably responsible for the more signi®cant effect of
melatonin on the organism, which approached that observed
when E. coli was incubated with vitamin E.
A Kendall agreement test was applied to determine
whether any parallelism exists between the hydrophobic
characteristics of E. coli and its adherence to nitrocellulose
®lters. The test revealed no association between the two
groups of results in any case, except when chloramphenicol + melatonin was applied in the EDDA-MH culture
broth. Therefore, the variations in surface hydrophobicity
and adherence to nitrocellulose ®lters must be determined
by surface characteristics of the bacteria that are interrelated
when E. coli is incubated in EDDA-MH.
Although the incubation of E. coli with or without
melatonin did not produce signi®cant differences in surface
hydrophobicity, a tendency was observed towards a reduction of the latter in the presence of melatonin. This
reduction was greater when E. coli was incubated with
another antioxidant, vitamin E. It is therefore concluded
that in E. coli, the antioxidant effect of vitamin E is re¯ected
as a signi®cant reduction in surface hydrophobicity. It is
noteworthy that E. coli did not respond in the same way to
melatonin, which in most situations is a powerful antioxidant (Daniels et al. 1995; Poeggeler et al. 1995). This could
be related to the genito-urinary origin of most of the E. coli
studied, as well as to the adaptation of the micro-organism to
the periodic presence of the hormone in the urine, bearing in
mind that a reduction in the surface hydrophobicity of
E. coli in the presence of endogenous melatonin (normally
eliminated in the urine) would mean that it would be
eliminated more readily from colonized urinary epithelia.
6REFERENCES
Criado, M.T., Pintor, M. and Ferreiros, C.M. (1993) Iron uptake by
Neisseria meningitidis. Research Microbiology 144, 77±82.
Daniels, W.M.U., Reiter, R.J., Melchiorri, D., Sewerynek, E., Pablos,
M.I. and Ortiz, G.G. (1995) Melatonin counteracts lipid peroxida-
tion induced by carbon tetrachloride but does not restore glucose 6
phosphatase activity. Journal of Pineal Research 19, 1±6.
Ellis, R.P. and Vorhies, M.W. (1976) Effect of supplemental dietary
vitamin E on the serologic response of swine to an Escherichia coli
bacterin. Journal of the American Veterinary Medicine Association 168,
231±232.
FerreiroÂs, C.M. and Criado, M.T. (1980) Adhesive properties
associated with the K99 antigen of Escherichia coli. IRCS Medical
Science 10, 917±918.
FerreiroÂs, C.M. and Criado, M.T. (1984) Expression of surface
hydrophobicity encoded by R-plasmids in Escherichia coli laboratory
strains. Archives of Microbiology 138, 191±194.
Friedman, A., Bartov, I. and Sklan, D. (1998) Humoral immune
response impairment following excess vitamin E nutrition in the
chick and turkey. Poultry Science 77, 956±962.
Fu, K., Hassett, D.J. and Cohen, M.S. (1989) Oxidant stress in
Neisseria gonorrhoeae: adaptation and effects on L-(+)lactate dehydrogenase activity. Infection and Immunity 57, 2173±2178.
Gibbons, R.J. (1996) Role of adhesion in microbial colonization of host
tissues: a contribution of oral microbiology. Journal of Dental
Research 75, 866±870.
Gibbons, R.J. and Hay, D.I. (1989) Adsorbed salivary acidic prolinerich proteins contribute to the adhesion of Streptococcus mutans JBP
to apatitic surfaces. Journal of Dental Research 68, 1303±1307.
Gibbons, R.J., Spinell, D.M. and Skobe, Z. (1976) Selective adherence
as a determinant of the host tropisms of certain indigenous and
pathogenic bacteria. Infection and Immunity 13, 238±246.
Hogan, J.S., Weiss, W.P., Smith, K.L., Todhunter, D.A., Schoenberger, P.S. and Williams, S.N. (1993) Vitamin E as an adjuvant in
an Escherichia coli J5 vaccine. Journal of Dairy Science 76, 401±407.
Lachica, R.V. and Zink, D.L. (1984) Plasmid-associated cell surface
charge and hydrophobicity of Yersinia enterocolitica. Infection and
Immunity 44, 540±543.
Li, J.C. and Xu, F. (1997) In¯uences of light-dark shifting on the immune
system, tumor growth and life span of rats, mice and fruit ¯ies as well as
on the counteraction of melatonin. Biological Signals 6, 77±89.
Poeggeler, B., Reiter, R.J., Hardeland, R., Sewerynek, E., Melchiorri, D.
and Barlowwalden, L.R. (1995) Melatonin, a mediator of electron
transfer and repair reactions, acts synergistically with the chain
breaking antioxidants ascorbate, trolox and glutathione. Neuroendocrinological Letters 17, 87±92.
Reiter, R.J., Tang, L., Garcia, J.J. and Munoz-Hoyos, A. (1997)
Pharmacological actions of melatonin in oxygen radical pathophysiology. Life Science 60, 2255±2271.
RodrõÂguez, A.B., Ortega, E., Lea, R.W. and Barriga, C. (1997)
Melatonin and the phagocytic process of heterophils from the ring
dove (Streptopelia risoria). Molecular Cell Biochemistry 168, 185±190.
Rosenberg, M., Gutnick, D. and Rosenberg, E. (1980) Adherence of
bacteria to hydrocarbons: a simple method for measuring cell-surface
hydrophobicity. FEMS Microbiology Letters 9, 29±33.
Stojiljkovic, I., Hwa, V., De Saint, M. et al. (1995) The Neisseria
meningitidis haemoglobin receptor: its role in iron utilization and
virulence. Molecular Microbiology 15, 531±541.
Uberos, J., Molina, A., LieÂbana, J., AugustõÂn, M.C. and MunÄoz, A.
(2000) The in¯uence of different concentrations of melatonin on the
cell-surface hydrophobic characteristics of Neisseria meningitidis.
Letters in Applied Microbiology 31, 294±298.
ã 2001 The Society for Applied Microbiology, Letters in Applied Microbiology, 32, 303±306